- Email: [email protected]
Isomerization and saturation of monounsaturated acyl moieties during catalytic hydrogenation influenced by their position in triacylglycerols
Chemistry and PhS,~tcsof Lip~ls 2i (1978) i87-t94
© FAsevier/Norm-HoltandScientific Publishers Ltd.
ISOMERIZATION AND SATURATION OF MONOUNSATURATED ACYL MOIETIES DURING CATALYTIC HYDR(K;ENATION INFLUENCED BY THEIR POSITION ~ T P d A C Y ~ L Y C E R O L S M.M. PAULOSE *, K.D. MUKHERJEE and I. RICHTEF Federal Center for Lipid Research, Piusallee, D-4400 M~,ster, GFR Rece~ve,d May 27, 1 9 7 7
accepted September I, 1977
1, 2, 3-Tri-(Z)-9,oetadeeenoylglyeerol(triolein) and 1, 2, 3-tri-(Z)-I3-docosenoylglycero! (trieruein) were partially hydrogenateA using a palladium catalyst. The unszturated acyl moieties at the 2-position of I, 2, 3-triaeylglye~ols wez~ reduced at a -slowerrate thar~those a~. the 1, 3-positions. The extent of geometrical isomerization ~mctly higher and the dc,ubte bonds w~e ~mewhat more scattered in the aeyl moieties at the 2-position than tho: ~.at 1- ar.d 3-positions.
L Introduction Little is known about the reactivity of unsaturated acyl moieties at various positions of triar ylglycerols during catalytic hydrogenation. Earlier ~nvestig:~tors who hydrogen~.~d mixtures of triacylglyeerols, ~ c h as r~adomlzed soybean oiJ a~ld o!;ve off, concluded that hydrogenation of an acyl moiety is not ;~qfluer~ccd by the position it cccupies in a triaeylglycerol [ I, 2]. More recently, however, the linoleoyl moiety of sunflower oil has been reported to hydrogenate faster ~r~ the 1,3-positions than in the 2position [3 ]. We subjected pure symmetrical triacylglycerols, 1,2,3-tri-(Z)-9-octadecenoyb glycerol (triolein)and 1,2,3-tri.(Z)-13-doeosenoylglycerol (trierucin), to partiM hydro~nation and analyzed the scyl moieties in the 2-position and I ~2,3-position~ of the triaeylglyeerols formed. We found that the extent of saturation as well as geometrical and positional isomerizafion of acyl moieties during hydroger~ation are i a d ~ d dependent on the position they occupy in triacylg!ycerols. IL Expegin~tal
i t Matenat
ecenoic acid (oleicacid:)~ A (Z)-13-doeosenoic acid (eruc~cacid) wej~ purcha~d from Nu.Chek-Prep, Elysian,Minn. 56028, USA. Olive oil ,ofD A 2 * Present ~ddress: Regional Re~..~rch Labo~ ~r/, Hydera bacl-9..Ind.. 187
188
g.M. Pauloee et al., Hydrogenation of monounsaturated acyl moieties
7 grade was purchased locally. Palladium chloride/barium sulfate contaiv±ng 10% Pd, porcine pancreas lipase as well as all other reagents of analyncal grade were obtained from E. Merck AG, D-6100 Darrnstadl, GFR. B, Methods
1,2,3-Tri-(Z)-9-octadecenoylglyeero! and 1,2,3-tri-(Z)- 13-doco~noylglycerol were prepared by esterification of the respective acids with glycerol usiagp-toiuenesulfonic acid as the catalyst, and purified by adsorption chromatography on silica gel. These compounds were found to be more th,'m 96% pure as determined by adsorption and argentation thin-layer chromatography, gas chromatography and reductive ozonolysis followed by gas chromatography (described later). Hydrogenations were carried out in a 20 ml screw-capped reaction tube provided with a magnetic stirrer. The tube ,:-as fitted with ~ teflon-tined septum, through which two stainless steel needles ~ e inserted to :mr,,e as inlet and outlet of hydrogen. Palladium chloride/barium sulfate, 10 nag, was suspended in 8 ml hexane in the reaction tube and reduced to the active catalyst by bubbling hydrogen through the reaction mixture under stirring at ambient temperature for 30 rain. The reaction tube was purged with nitrogen and the tri~wylglycerol, 100 mg, dissolved in 2 ml hexane, was added to the catalyst suspension. Hydrogenation was started by bubbling hydrogen through the reaction mixture under stirring at 25°C. Partially hydrogenated samples were withdrawn from the re~tioa mixture after 15 min and 60 min and centrifuged to separate the catalyst. The triacylglycerols used as starting materials, as well as the partially hydrogenated products, were analyzed as follows. An aliquot of the triacylglycerols was converted to methyl esters by transersterifi,:ation. The methyl esters were analyzed by. gas chromatography us_tag flame ionization detectors. The separations were carried o,st on a column, 6 f t b y ~ in., packed with 10% EGSS-X on Gas-Chrom P, 100-120 mesh (Applied Science Laboratories Inc., State College, Pa. 16801, USA). The proportions of the various components in each mixture were ca/culated as percentage area of the r~spective peak, measured by triangulation. The remaining part of the methyl esters was fractionated into (Z)-and (E~ isomers by argentation chromatography [4] on a layer of silica gel contorting I ~ silver nitrate. The plates were developed twice with hexane-diethyl ether (90:10). The methyl esters were visualized by spraying with 0.1% ethanolic 2', 7"dichlorofluorescein and extracted with water-saturated diethyl ether. A definite amount of methyl eicosaneate was added as internal standard to aliquots of the (Z)-and(E) isomers and the ratio of these isomers was determined by gas chromatography, as described above. An aliquot of the methyl esters, ca. 100 #g, was dissolved in pent~e ~ d ozonized at -70°C using a Supelco micro-ozonizer [5,6]. The reaction products we re evaporated to dryneas at room temperature, diss~ved m ~ tAearbon disul-
P&M. Pauiose et aL , Hydrogeraation o f monounse tura t ed acy l m gie ties
189
file, and the ozonides reduced using I - 3 mg of tfiphenylphesphLne. After- staadmg at room temperature for 30 redn, the mixt-are of aldehydes and ",aldehydeesters was ~ y z e d by ~ chromatograp.ay using a column, 6 ft b:~ ~ in., packed wi~h 10% OV-I 7on Gas-Chrom Q, 8 0 - i 90 mesh Q~'GA, D-4OO0 D~sseldorL (~FR), coupled with mother column, 2 ft by -~ ia~, packed wi~ 3% OV-225 on ~e!coport (Supeleo, Inc., ~ l e f o n t e , Pa. 16823, USA). The temperature was p~ogram° med from 50eC to 270°C, 5°C/mm [7,8]. Identification of fragments was by e homologom series of aldehydes and by ozonolysis of monounsaturated methyl esters ofknown positional configuration. Quantitative results were obtained from peak ar~as, measured by triangulation; response factors were applied to correct for lack of xesponse m the flame ion~ation detector [~y carb axyl and carbonyl carbon atoms [91 A part of each sample of tfiaeylglycerol was hydrolyzed with porcine p~-lcreatic lipase [ 10]. The 2-acylgly~rols formed were isolated by preparative thhl-layer chromatography on silica gel G eontahuing 8% boric acid [ 11 ] with he×ane-dieCr~yl ether-aceti!c acid (50:50:1) as the developing solvent. As described above, the 2-aeylglyce~rolswere converted to methyl esters, which were analyzed by gas chromatography. The methyl esters were then fractionated into (Z)- and (E)qmmers by argentation thin-layer chromatography. The fractions obtained we re subjected to reductive ozonolysis and the fragments were analyzed by gas chromztography. HI. Results and discussion
In order to assess the extent to which acyl migration might occur during partial hydrogenation of 1,2,3-tri-(Z)-9-~tadecengylglycerol and 1,2,3-triqZ~ 13-docosenoylslycerol, the triacylglycerols of olive off were partially hydrogenated under similar conditions. As criteria for acyl migration, we followed the levels of he×aTable 1 Fatty acid composition (%) of o?~'¢e oil and partially hydrogenated products Acyl moieties * 16:0
16:1
18:0
18:1
18:2
18:3
11.1 0.6
1.0 0.8
2.8 0.1
74.9 82.1
9.3 15.6
0.9 0.8
0.8 0.7
14.6 11.3
72.8 85.8
0.3 0.9
0.2 0.2
0 min hydrogenartion
(starting material) 1,2,3-Triaey~yeerol 2-Acylglycerol 15 min hydrogenation
1,2,3-Triacylglycerol 2-Acylgtycerot
11.3 1.1
* Number of carbon atoms : number of double bonds.
190
M.M. Pauiose et ~;., Hydrogenation o f m,rnounsatv~,ated a~yl m o i e ~
decanoic acid and hexadecanoic acids in the tric,ytglycerois as well as in the corresponding 2-acylgtycerols, derived by lipase hydrolysis befi~re and after hydrogenation. The ~sults given in table I show that after the reduction of ca. 22%of the double bonds, randomization between 2- and 1, of the triaeylglycerols occurred to an extent of approximately 15% only, The extent:of randomization was estimated as follows. A complete r a n d o ~ t i o n would vidd. at a level of 12.1% of fatty acids with 16 carbon atoms in t erols, 12.113 = 4.03% of these acids in the 2-acyiglyczrols. But the actual amount of these fatty acids present in the 2-acylglycerols before hydrogenation is 1,4%. Therefore, in the ease of complete randomization the content of these" fatty acids would increase by 4.03 1.4 = 2.63%. After hydro~ceeation, the content of fatty acids with 16 carbon atoms increased by 1.8 - ! .4 = 0.4% in the 2-acylglyeerols, which implies randomization to an extent of approximately (0.4t2.63) X 100 = ca. 15%. Since aeyl migration is a prerequisite for randomization, we assume that the overall extent of acyl migration was also of the same order of magnitude. The fatty acid compositic.n of the partially hydrogenated products from 1,2,3tri-(Z)-9-octadecenoylglycerol and 12,3-tri,(Z)-13-doeosenoylglycerol and of the 2-acylglycerols derived therefrom by lipase hydrolysis is given in tables 2 and 3. Apparent!y, even after 60 rain much less double l:~nds were reduced in each of these triacylglycerols, compared to partially hydrogenated olive off (table 1). We ~ssume, therefore, that acyl migration between 2- and 1,3-positions, if any, occurred to an extent less than 15% during the hydrogenation of 1,2,34ri(Z)-9.octa. decenoylgJycerol and 1,2,3
Table 2 Fatty acid composition (%) of partially hydrogenated products derived from 1,2,3-tri-(Z)-9octadecenoylglycer¢l Acyl moieties * 18:0
18:1 isomers (Z) + (E)
(Z)**
(E)**
3.7 3.2
96.3 96.8
95.1 93.0
4,9 7.0
4.2 2.3
95.8 97.2
94.5 92.6
5.5 7.4
15 rain hydrogenation
1,2,3*Triacylglycerol 2-Acylglycerol 60 rain hydrogenation
1,2,3oTfiacylglycerol 2-Acylglyccrol
* Number of carbon atom," : number of double bonds. ** % total 18:1.
teL.M. Pat,lose e~~aL, Hydrogenation ~f moneunsararated acyl moie~ges
I91
Table 3 Fatty aeki composition (%) of partially hydrogenated products derived from 1,2,3-~rb(Z)-i3de~o~noylglycero| Aeyi moieties * 22:0 22:1 isomers (Z) + (E) (Z)**
(E)**
15 min hydrogenation
1,2,3-T" r t a ~ , ~ o l 2-Aeylgly~tol
4.4 3.1
95.6 96.9
88.2 84.2
t 1.8 t5.8
6.6 3.3
93.2 96.5
82.9 75.7
17.1 24.3
60 rain hydrogenation
1,2,3-Tria~lglyo~rol 2-Acylglyeerol
* Number of carbon ~toms : number of double bonds. ** % total 22:1.
in the triacylglycerols compared to that in the corresponding 2-acylglyceroiso it is thus evident that the ra~e of hydrogenation is lower ~ d the rate of geometrical isomerizs;:ion is higher in th, 2-position of triacylglyeerol~ than in the 1 #-positions. It is interesting to note flint at nearly the same level of reduction, 4.4% and 4~2% tmturates, reSl~ctively, rite content of (E~-isomers in the partiMly hydrogenated product from 1,2~3-tfl-(Z)-I 3Moc,osenoyiglyeerol as well as in the 2-acylgiyceroL,, derived therefrom by l i p ~ Mdrolysis was m6re than twice as high as in the corresponding products obtained from 1,2,3-tri-(Z)-9-octade~noylglyceroI TI'~ composition of (Z)~mld (E)-isomers in partially hydrogenated products derived from 1,2~34ri~(Z)-9-oetsdeeenoylglyc, ml and 1,2~3-tfi-(Z)- 13-docosenoy]glycerol, which was determined by reduetive ozonolysis and gas chromatography, is siren/n tabl~ 4 and 5, reslt~etively. The resulta given in table 4 show that in the partially hydrogenated products derived from 1,2,3-tri-(Z)-9-cctade~noylgly-erol after 15 and 60 min, corresponding to 3.7 mad 4.2% reduetit % respectively, the proportion of the (Z)-9oisomer in the triaeylglyeeroh is almost ~ high as in the starting material, whereas the content of this isomer in the 2.aeylglyeerols, obtained by lipase hydrolysis, is considerably l o ~ r . It is therefore evident ~lttat the aeyl moieties in 1,3-posifiom of the triacy!o glycerols are almost exclusively- comprised of (Z)-9-octadeeenoate, whereas some scattering of double bonds oe:urred p~dominanfly in the 2-posiuon. contrast, ~ e double bonds in the 0~isomers derived from 1,2,3-tri-(Z)-9oetadecen~lglycerol are widely scattered; however, there is no significan~ difference in the ?attem of distribution of (E}isomers in the 2- and 1,3opositions of tl~ triacylglyeer~ ~L~.It is interesting to note th~tt, in ~ e aeyI moieties, both in 2o arid 1,3-positions oflt.;~eylgiycerc,L~ the migration of double bo~ds i~ preferen~a~Jy
2-Acy~ycerol . . . .
1,2,3-Tria~lglycerol
60 min hydr°genation
2-Acylglycerol
1,2~-Tnacyiglycerol
15 rain hydrogenation
~)
(E)
(Z)
(Z) (E) (Z) rE)
0.8
2.6 1,4
2
1.3 1.1 2.5
0.1 1.0 2.2 3.7
3
1.0 0.7 1.9
tr 0.9 1.3 3.7
4
tr 1.1 0.1 0.8
~ 0.4 3.2 0.9
5
tr 1.6 5.7 4.8
,C.C, 1,7 0.9 6.0
6
7
8
tr 2.8 1.2 1.8
1.8 i5.0 9.0 24.4
~).2 1.8 2.2 27.8 05 3.7 1.6 18.9
Position of double bonds (A)
97.8 63.2 81.6 59.5
96.5 55.3 ~2.5 55.6
9
0.2 3.1 0.8 2.1
0.1 2.1 ;.6 2.3
10
0.2 8.7 tr 1.2
0.5 7.7 0.t 2.8
!!
tr 0.4 tr tr
tr 0.4 1.0 1,1
12
tr
tr 6.8
tr tr 0.2 1.5
i3
tr
0.4
~ ~.,~ 0,2 0.6
14
tr
0.8
0.1 tr
tr
15
Table 4 Composition (%) of (Z)- and (E)-isomers in partially hydrogenated products derived from 1,2,3-tri-(Z~9-octadecenoylglycerol
to
M.M. PapOose et aL, Hydrogenation o f mono~nsamrated acyZ moieties
t 93
Table 5 Composition (%) of (Z)- and @;)-~omers Ln partially hydrogenated products derived f~om ~,2~3-
*~2-(Z)-13-doe~senoylglyc~ol Position of double bonds (~) 6
7
8
9
10
11
12
13
1,4
15
~5
I7
~) ~) ~) ~)
0,1 0.1 0.1 0.4
0.3 0,2 ~2 0,8
~ 0,7 0.1 0.7
~ 0.6 0.1 0.6
3,1 0.7 0.2 1.0
0.4 0.9 0,1 2.1
0,9 15.5 1.5 19.2
97.2 59.5 93.4 51.7
0,2 18.5 0,7 17,2
0,6 2.4 3,4 4,0
01 0~ 03 t,~
0.5
~) ~) ~) ~)
0.1 0.1 0,1 0.2
0,1 0.2 0.2 0.8
0. i 0.3 0,1 0.6
0-1 0.4 0.2 0.6
0.1 0.6 0.2 0,9
0.2 1.6 0.1 1.9
1.6 15.8 1,2 18.8
~.0 54.3 93.7 54.4
0.9 I8.4 0,6 16,8
37 5,9 3.4 '3,0
0.1 1.8 ),t ~.5
Oi. 0.6 0. t 0~5
15 re.in hydrogenation 1,2,3..Triaeylglyeerol
2-Acylglyeerol 60 mm hydrogenati¢ n
1,2,3-Triacylglyeerol 2-Acylglycerol
directed towards the 8-position rather than the l 0- position. The results given in table 5 show that in the partially hydrogenated p~'oduct derived from 1,2,3-tri-(Z)-13-docosenoylglycerot after i 5 rain, correspondi~tg to 4.4% reduction, the proportion of (Z)-13-isomer in the triacylglycerois is high, whereas the proportion of this isomer in the 2-acylglycero] obtained by Iip;~se hydrolysis, is shghfly l)wer. However, in the product obtained after 63 ~ir~ of hydrogenation, corresponding to 6.6% reduction, such differences are n o f ~uTC. It se,;ms that in the part_~alhydrogenation of 1,2,3-tri-(Z)-I 3-docosenoylglycerot there are no pronounced differences in the scattering of double bonds in the (Z)isomers, regardless of their position in the triacylglycerols. The (E)-isom~rs in the triacylglycerols derived from 1,2,3-tfi-(Z)- t 3-docosenoyL glycercl have the double bonds as widely scattered as in the corresp(,nding prod~cts derived from 1,2,3-tri.(Z)-9-octadecenoylglycerol. Again, there is oo significant difference in the pattern of distribution of (E)-isomers in the 2- and 1,3-po:.;it!o~s of the triaeylglycerols. In contrast to the partial hydrogenation of 1,2,3-tfi-(Z}9 oetadeeenoylglycerol, in whieh the double bonds preferentially migrated into the vieinal position towards flie methyl end, in the hydrogenation of 1,2,3-tri-(Z)-13doeosenoylglyeero!, the neigt boufing positions of the original do~ble bond are occupied to an almost equal extent.
Aeknov4edgement Financial support provided by 'Deutscher Akadem~cher Austauschdie~st (DAAD)' is gratefully acknowledged.
l~qeamces
4t3
[31 F . L u k ~ , E . ~
57
M: ~ j d ~ M . r . ~ i ~ ~ ~ ~ ! O ~
[7] A.E.joh~mtd HJ. [ ~ ,
J~Oil~!~.4Jgi(|~9$
[t~l M.YurkowskiamlH.B r ~ , ] ~ _ . ~ . ~ ~ 1 2 5 ( I ~ ) 5 5
© FAsevier/Norm-HoltandScientific Publishers Ltd.
ISOMERIZATION AND SATURATION OF MONOUNSATURATED ACYL MOIETIES DURING CATALYTIC HYDR(K;ENATION INFLUENCED BY THEIR POSITION ~ T P d A C Y ~ L Y C E R O L S M.M. PAULOSE *, K.D. MUKHERJEE and I. RICHTEF Federal Center for Lipid Research, Piusallee, D-4400 M~,ster, GFR Rece~ve,d May 27, 1 9 7 7
accepted September I, 1977
1, 2, 3-Tri-(Z)-9,oetadeeenoylglyeerol(triolein) and 1, 2, 3-tri-(Z)-I3-docosenoylglycero! (trieruein) were partially hydrogenateA using a palladium catalyst. The unszturated acyl moieties at the 2-position of I, 2, 3-triaeylglye~ols wez~ reduced at a -slowerrate thar~those a~. the 1, 3-positions. The extent of geometrical isomerization ~mctly higher and the dc,ubte bonds w~e ~mewhat more scattered in the aeyl moieties at the 2-position than tho: ~.at 1- ar.d 3-positions.
L Introduction Little is known about the reactivity of unsaturated acyl moieties at various positions of triar ylglycerols during catalytic hydrogenation. Earlier ~nvestig:~tors who hydrogen~.~d mixtures of triacylglyeerols, ~ c h as r~adomlzed soybean oiJ a~ld o!;ve off, concluded that hydrogenation of an acyl moiety is not ;~qfluer~ccd by the position it cccupies in a triaeylglycerol [ I, 2]. More recently, however, the linoleoyl moiety of sunflower oil has been reported to hydrogenate faster ~r~ the 1,3-positions than in the 2position [3 ]. We subjected pure symmetrical triacylglycerols, 1,2,3-tri-(Z)-9-octadecenoyb glycerol (triolein)and 1,2,3-tri.(Z)-13-doeosenoylglycerol (trierucin), to partiM hydro~nation and analyzed the scyl moieties in the 2-position and I ~2,3-position~ of the triaeylglyeerols formed. We found that the extent of saturation as well as geometrical and positional isomerizafion of acyl moieties during hydroger~ation are i a d ~ d dependent on the position they occupy in triacylg!ycerols. IL Expegin~tal
i t Matenat
ecenoic acid (oleicacid:)~ A (Z)-13-doeosenoic acid (eruc~cacid) wej~ purcha~d from Nu.Chek-Prep, Elysian,Minn. 56028, USA. Olive oil ,ofD A 2 * Present ~ddress: Regional Re~..~rch Labo~ ~r/, Hydera bacl-9..Ind.. 187
188
g.M. Pauloee et al., Hydrogenation of monounsaturated acyl moieties
7 grade was purchased locally. Palladium chloride/barium sulfate contaiv±ng 10% Pd, porcine pancreas lipase as well as all other reagents of analyncal grade were obtained from E. Merck AG, D-6100 Darrnstadl, GFR. B, Methods
1,2,3-Tri-(Z)-9-octadecenoylglyeero! and 1,2,3-tri-(Z)- 13-doco~noylglycerol were prepared by esterification of the respective acids with glycerol usiagp-toiuenesulfonic acid as the catalyst, and purified by adsorption chromatography on silica gel. These compounds were found to be more th,'m 96% pure as determined by adsorption and argentation thin-layer chromatography, gas chromatography and reductive ozonolysis followed by gas chromatography (described later). Hydrogenations were carried out in a 20 ml screw-capped reaction tube provided with a magnetic stirrer. The tube ,:-as fitted with ~ teflon-tined septum, through which two stainless steel needles ~ e inserted to :mr,,e as inlet and outlet of hydrogen. Palladium chloride/barium sulfate, 10 nag, was suspended in 8 ml hexane in the reaction tube and reduced to the active catalyst by bubbling hydrogen through the reaction mixture under stirring at ambient temperature for 30 rain. The reaction tube was purged with nitrogen and the tri~wylglycerol, 100 mg, dissolved in 2 ml hexane, was added to the catalyst suspension. Hydrogenation was started by bubbling hydrogen through the reaction mixture under stirring at 25°C. Partially hydrogenated samples were withdrawn from the re~tioa mixture after 15 min and 60 min and centrifuged to separate the catalyst. The triacylglycerols used as starting materials, as well as the partially hydrogenated products, were analyzed as follows. An aliquot of the triacylglycerols was converted to methyl esters by transersterifi,:ation. The methyl esters were analyzed by. gas chromatography us_tag flame ionization detectors. The separations were carried o,st on a column, 6 f t b y ~ in., packed with 10% EGSS-X on Gas-Chrom P, 100-120 mesh (Applied Science Laboratories Inc., State College, Pa. 16801, USA). The proportions of the various components in each mixture were ca/culated as percentage area of the r~spective peak, measured by triangulation. The remaining part of the methyl esters was fractionated into (Z)-and (E~ isomers by argentation chromatography [4] on a layer of silica gel contorting I ~ silver nitrate. The plates were developed twice with hexane-diethyl ether (90:10). The methyl esters were visualized by spraying with 0.1% ethanolic 2', 7"dichlorofluorescein and extracted with water-saturated diethyl ether. A definite amount of methyl eicosaneate was added as internal standard to aliquots of the (Z)-and(E) isomers and the ratio of these isomers was determined by gas chromatography, as described above. An aliquot of the methyl esters, ca. 100 #g, was dissolved in pent~e ~ d ozonized at -70°C using a Supelco micro-ozonizer [5,6]. The reaction products we re evaporated to dryneas at room temperature, diss~ved m ~ tAearbon disul-
P&M. Pauiose et aL , Hydrogeraation o f monounse tura t ed acy l m gie ties
189
file, and the ozonides reduced using I - 3 mg of tfiphenylphesphLne. After- staadmg at room temperature for 30 redn, the mixt-are of aldehydes and ",aldehydeesters was ~ y z e d by ~ chromatograp.ay using a column, 6 ft b:~ ~ in., packed wi~h 10% OV-I 7on Gas-Chrom Q, 8 0 - i 90 mesh Q~'GA, D-4OO0 D~sseldorL (~FR), coupled with mother column, 2 ft by -~ ia~, packed wi~ 3% OV-225 on ~e!coport (Supeleo, Inc., ~ l e f o n t e , Pa. 16823, USA). The temperature was p~ogram° med from 50eC to 270°C, 5°C/mm [7,8]. Identification of fragments was by e homologom series of aldehydes and by ozonolysis of monounsaturated methyl esters ofknown positional configuration. Quantitative results were obtained from peak ar~as, measured by triangulation; response factors were applied to correct for lack of xesponse m the flame ion~ation detector [~y carb axyl and carbonyl carbon atoms [91 A part of each sample of tfiaeylglycerol was hydrolyzed with porcine p~-lcreatic lipase [ 10]. The 2-acylgly~rols formed were isolated by preparative thhl-layer chromatography on silica gel G eontahuing 8% boric acid [ 11 ] with he×ane-dieCr~yl ether-aceti!c acid (50:50:1) as the developing solvent. As described above, the 2-aeylglyce~rolswere converted to methyl esters, which were analyzed by gas chromatography. The methyl esters were then fractionated into (Z)- and (E)qmmers by argentation thin-layer chromatography. The fractions obtained we re subjected to reductive ozonolysis and the fragments were analyzed by gas chromztography. HI. Results and discussion
In order to assess the extent to which acyl migration might occur during partial hydrogenation of 1,2,3-tri-(Z)-9-~tadecengylglycerol and 1,2,3-triqZ~ 13-docosenoylslycerol, the triacylglycerols of olive off were partially hydrogenated under similar conditions. As criteria for acyl migration, we followed the levels of he×aTable 1 Fatty acid composition (%) of o?~'¢e oil and partially hydrogenated products Acyl moieties * 16:0
16:1
18:0
18:1
18:2
18:3
11.1 0.6
1.0 0.8
2.8 0.1
74.9 82.1
9.3 15.6
0.9 0.8
0.8 0.7
14.6 11.3
72.8 85.8
0.3 0.9
0.2 0.2
0 min hydrogenartion
(starting material) 1,2,3-Triaey~yeerol 2-Acylglycerol 15 min hydrogenation
1,2,3-Triacylglycerol 2-Acylgtycerot
11.3 1.1
* Number of carbon atoms : number of double bonds.
190
M.M. Pauiose et ~;., Hydrogenation o f m,rnounsatv~,ated a~yl m o i e ~
decanoic acid and hexadecanoic acids in the tric,ytglycerois as well as in the corresponding 2-acylgtycerols, derived by lipase hydrolysis befi~re and after hydrogenation. The ~sults given in table I show that after the reduction of ca. 22%of the double bonds, randomization between 2- and 1, of the triaeylglycerols occurred to an extent of approximately 15% only, The extent:of randomization was estimated as follows. A complete r a n d o ~ t i o n would vidd. at a level of 12.1% of fatty acids with 16 carbon atoms in t erols, 12.113 = 4.03% of these acids in the 2-acyiglyczrols. But the actual amount of these fatty acids present in the 2-acylglycerols before hydrogenation is 1,4%. Therefore, in the ease of complete randomization the content of these" fatty acids would increase by 4.03 1.4 = 2.63%. After hydro~ceeation, the content of fatty acids with 16 carbon atoms increased by 1.8 - ! .4 = 0.4% in the 2-acylglyeerols, which implies randomization to an extent of approximately (0.4t2.63) X 100 = ca. 15%. Since aeyl migration is a prerequisite for randomization, we assume that the overall extent of acyl migration was also of the same order of magnitude. The fatty acid compositic.n of the partially hydrogenated products from 1,2,3tri-(Z)-9-octadecenoylglycerol and 12,3-tri,(Z)-13-doeosenoylglycerol and of the 2-acylglycerols derived therefrom by lipase hydrolysis is given in tables 2 and 3. Apparent!y, even after 60 rain much less double l:~nds were reduced in each of these triacylglycerols, compared to partially hydrogenated olive off (table 1). We ~ssume, therefore, that acyl migration between 2- and 1,3-positions, if any, occurred to an extent less than 15% during the hydrogenation of 1,2,34ri(Z)-9.octa. decenoylgJycerol and 1,2,3
Table 2 Fatty acid composition (%) of partially hydrogenated products derived from 1,2,3-tri-(Z)-9octadecenoylglycer¢l Acyl moieties * 18:0
18:1 isomers (Z) + (E)
(Z)**
(E)**
3.7 3.2
96.3 96.8
95.1 93.0
4,9 7.0
4.2 2.3
95.8 97.2
94.5 92.6
5.5 7.4
15 rain hydrogenation
1,2,3*Triacylglycerol 2-Acylglycerol 60 rain hydrogenation
1,2,3oTfiacylglycerol 2-Acylglyccrol
* Number of carbon atom," : number of double bonds. ** % total 18:1.
teL.M. Pat,lose e~~aL, Hydrogenation ~f moneunsararated acyl moie~ges
I91
Table 3 Fatty aeki composition (%) of partially hydrogenated products derived from 1,2,3-~rb(Z)-i3de~o~noylglycero| Aeyi moieties * 22:0 22:1 isomers (Z) + (E) (Z)**
(E)**
15 min hydrogenation
1,2,3-T" r t a ~ , ~ o l 2-Aeylgly~tol
4.4 3.1
95.6 96.9
88.2 84.2
t 1.8 t5.8
6.6 3.3
93.2 96.5
82.9 75.7
17.1 24.3
60 rain hydrogenation
1,2,3-Tria~lglyo~rol 2-Acylglyeerol
* Number of carbon ~toms : number of double bonds. ** % total 22:1.
in the triacylglycerols compared to that in the corresponding 2-acylglyceroiso it is thus evident that the ra~e of hydrogenation is lower ~ d the rate of geometrical isomerizs;:ion is higher in th, 2-position of triacylglyeerol~ than in the 1 #-positions. It is interesting to note flint at nearly the same level of reduction, 4.4% and 4~2% tmturates, reSl~ctively, rite content of (E~-isomers in the partiMly hydrogenated product from 1,2~3-tfl-(Z)-I 3Moc,osenoyiglyeerol as well as in the 2-acylgiyceroL,, derived therefrom by l i p ~ Mdrolysis was m6re than twice as high as in the corresponding products obtained from 1,2,3-tri-(Z)-9-octade~noylglyceroI TI'~ composition of (Z)~mld (E)-isomers in partially hydrogenated products derived from 1,2~34ri~(Z)-9-oetsdeeenoylglyc, ml and 1,2~3-tfi-(Z)- 13-docosenoy]glycerol, which was determined by reduetive ozonolysis and gas chromatography, is siren/n tabl~ 4 and 5, reslt~etively. The resulta given in table 4 show that in the partially hydrogenated products derived from 1,2,3-tri-(Z)-9-cctade~noylgly-erol after 15 and 60 min, corresponding to 3.7 mad 4.2% reduetit % respectively, the proportion of the (Z)-9oisomer in the triaeylglyeeroh is almost ~ high as in the starting material, whereas the content of this isomer in the 2.aeylglyeerols, obtained by lipase hydrolysis, is considerably l o ~ r . It is therefore evident ~lttat the aeyl moieties in 1,3-posifiom of the triacy!o glycerols are almost exclusively- comprised of (Z)-9-octadeeenoate, whereas some scattering of double bonds oe:urred p~dominanfly in the 2-posiuon. contrast, ~ e double bonds in the 0~isomers derived from 1,2,3-tri-(Z)-9oetadecen~lglycerol are widely scattered; however, there is no significan~ difference in the ?attem of distribution of (E}isomers in the 2- and 1,3opositions of tl~ triacylglyeer~ ~L~.It is interesting to note th~tt, in ~ e aeyI moieties, both in 2o arid 1,3-positions oflt.;~eylgiycerc,L~ the migration of double bo~ds i~ preferen~a~Jy
2-Acy~ycerol . . . .
1,2,3-Tria~lglycerol
60 min hydr°genation
2-Acylglycerol
1,2~-Tnacyiglycerol
15 rain hydrogenation
~)
(E)
(Z)
(Z) (E) (Z) rE)
0.8
2.6 1,4
2
1.3 1.1 2.5
0.1 1.0 2.2 3.7
3
1.0 0.7 1.9
tr 0.9 1.3 3.7
4
tr 1.1 0.1 0.8
~ 0.4 3.2 0.9
5
tr 1.6 5.7 4.8
,C.C, 1,7 0.9 6.0
6
7
8
tr 2.8 1.2 1.8
1.8 i5.0 9.0 24.4
~).2 1.8 2.2 27.8 05 3.7 1.6 18.9
Position of double bonds (A)
97.8 63.2 81.6 59.5
96.5 55.3 ~2.5 55.6
9
0.2 3.1 0.8 2.1
0.1 2.1 ;.6 2.3
10
0.2 8.7 tr 1.2
0.5 7.7 0.t 2.8
!!
tr 0.4 tr tr
tr 0.4 1.0 1,1
12
tr
tr 6.8
tr tr 0.2 1.5
i3
tr
0.4
~ ~.,~ 0,2 0.6
14
tr
0.8
0.1 tr
tr
15
Table 4 Composition (%) of (Z)- and (E)-isomers in partially hydrogenated products derived from 1,2,3-tri-(Z~9-octadecenoylglycerol
to
M.M. PapOose et aL, Hydrogenation o f mono~nsamrated acyZ moieties
t 93
Table 5 Composition (%) of (Z)- and @;)-~omers Ln partially hydrogenated products derived f~om ~,2~3-
*~2-(Z)-13-doe~senoylglyc~ol Position of double bonds (~) 6
7
8
9
10
11
12
13
1,4
15
~5
I7
~) ~) ~) ~)
0,1 0.1 0.1 0.4
0.3 0,2 ~2 0,8
~ 0,7 0.1 0.7
~ 0.6 0.1 0.6
3,1 0.7 0.2 1.0
0.4 0.9 0,1 2.1
0,9 15.5 1.5 19.2
97.2 59.5 93.4 51.7
0,2 18.5 0,7 17,2
0,6 2.4 3,4 4,0
01 0~ 03 t,~
0.5
~) ~) ~) ~)
0.1 0.1 0,1 0.2
0,1 0.2 0.2 0.8
0. i 0.3 0,1 0.6
0-1 0.4 0.2 0.6
0.1 0.6 0.2 0,9
0.2 1.6 0.1 1.9
1.6 15.8 1,2 18.8
~.0 54.3 93.7 54.4
0.9 I8.4 0,6 16,8
37 5,9 3.4 '3,0
0.1 1.8 ),t ~.5
Oi. 0.6 0. t 0~5
15 re.in hydrogenation 1,2,3..Triaeylglyeerol
2-Acylglyeerol 60 mm hydrogenati¢ n
1,2,3-Triacylglyeerol 2-Acylglycerol
directed towards the 8-position rather than the l 0- position. The results given in table 5 show that in the partially hydrogenated p~'oduct derived from 1,2,3-tri-(Z)-13-docosenoylglycerot after i 5 rain, correspondi~tg to 4.4% reduction, the proportion of (Z)-13-isomer in the triacylglycerois is high, whereas the proportion of this isomer in the 2-acylglycero] obtained by Iip;~se hydrolysis, is shghfly l)wer. However, in the product obtained after 63 ~ir~ of hydrogenation, corresponding to 6.6% reduction, such differences are n o f ~uTC. It se,;ms that in the part_~alhydrogenation of 1,2,3-tri-(Z)-I 3-docosenoylglycerot there are no pronounced differences in the scattering of double bonds in the (Z)isomers, regardless of their position in the triacylglycerols. The (E)-isom~rs in the triacylglycerols derived from 1,2,3-tfi-(Z)- t 3-docosenoyL glycercl have the double bonds as widely scattered as in the corresp(,nding prod~cts derived from 1,2,3-tri.(Z)-9-octadecenoylglycerol. Again, there is oo significant difference in the pattern of distribution of (E)-isomers in the 2- and 1,3-po:.;it!o~s of the triaeylglycerols. In contrast to the partial hydrogenation of 1,2,3-tfi-(Z}9 oetadeeenoylglycerol, in whieh the double bonds preferentially migrated into the vieinal position towards flie methyl end, in the hydrogenation of 1,2,3-tri-(Z)-13doeosenoylglyeero!, the neigt boufing positions of the original do~ble bond are occupied to an almost equal extent.
Aeknov4edgement Financial support provided by 'Deutscher Akadem~cher Austauschdie~st (DAAD)' is gratefully acknowledged.
l~qeamces
4t3
[31 F . L u k ~ , E . ~
57
M: ~ j d ~ M . r . ~ i ~ ~ ~ ~ ! O ~
[7] A.E.joh~mtd HJ. [ ~ ,
J~Oil~!~.4Jgi(|~9$
[t~l M.YurkowskiamlH.B r ~ , ] ~ _ . ~ . ~ ~ 1 2 5 ( I ~ ) 5 5